Students intentionally degrade their writing to signal human authorship amid AI concerns

A student deliberately left a newspaper title unitalicized in a final paper—despite knowing the correct formatting—as an intentional signal to the professor that a human, not ChatGPT, wrote it.

Writers are shifting their goal from writing well to writing in ways that signal authenticity to humans. The author reports cutting em-dashes from writing completely a year ago, viewing them as an obvious fingerprint of LLMs (large language models that generate text), and sometimes degrades writing quality in their own blog for the same purpose.

The underlying anxiety is that the new measure of quality in writing has become the appearance of authenticity rather than authenticity itself, flattening the actual conversation about whether people are developing a voice and doing intellectual work into a binary judgment of human or machine origin.